Verbio Automated Speech Recognition (ASR) is VERBIO's speech recognition engine for telephone and multimedia environments.
Verbio Automated Speech Recognition is the technology that can automatically convert, speech to text independently of who is speaking, allowing the system to "understand" or interpret the content of a voiceover regardless of the voice.
Initially oriented towards telephone environments, it currently has a wide variety of both telehone and multimedia applications where voice recognition is the main interactive channel. The system recognizes specific words or sets of words that are said among a group of options (called a grammar). The furthest application of this is Verbio's natural language recognition that interprets speech as it occurs, without grammars.
The high recognition rates of Verbio's speech recognition are due to its capacity for adaptation in every environment; not only in terms of grammar or options, but also in terms of a great number of existing acoustic models.
In embedded, centralized or local environments, Verbio ASR is an essential tool in interactive applications between users and automatic or voice controled systems, whether they are IVR, voice portals, home automation systems, call centers, security systems, industrial applications, voice navigation or mobile devices in general.
In the cases in which there is an environment with a broad vocabulary, Verbio's solution is VoxPopuli, the transcription engine based on statistical patterns (SLM).
Verbio Vox Populi's objective is the transcription of spontaneous dialogues and providing a desk dictation system. Both objectives are unique and require different approaches.
Verbio Vox Populi, through its three axes (Acoustic Model, Natural Language Model and Transcription Engine) that adapt to the environment, is able to transform audio into text, with high reliability in very unfavorable situations, for example: noisy environments, large distances, high number of voices and various different languages,etc.
Verbio VoxPopuli is a system that functions regardless of the announcer since it is able to understand any person with great accuracy thanks to a design that was created from a database of thousands of voices of people per language, selected according to strict geographic and demographic criteria.
Verbio VoxPopuli works in real time. It is faster in dictation environments, and in transcription environments time increases in accordance with the difficulty in understanding (new words, ambience noise, etc.).